Machine Learning in Action: Securing IAM API by Risk Authentication Decision Engine

摘要

In this work we show the implementation of practical machine learning (ML) techniques into a DevOps ecosystem. Our system Risk Authentication/Assessment Decision Engine (RADE) is a ML model that uses digital identity’s attributes such as IP address, browser user agent, and user activity to estimate risk level of each single authentication attempt. The model is part of a risk evaluation application programming interface (API). We have developed a risk classifier model for identifying user risk level and validate our system on a sample of login data attributes.